Mathematics of Data Science VIrtual Lecture Series: Data-Dependent Distances for Low Sample Learning

Abstract

Approaches to unsupervised clustering and semisupervised learning with data-dependent distances are proposed. By considering metrics derived from data-driven graphs, robustness to noise and class geometry is achieved. The proposed algorithms enjoy theoretical guarantees on flexible data models, and also have quasilinear computational complexity in the number of data points. Applications to remote sensing images and biological networks will be shown, demonstrating the practical applicability of our methods to real problems in machine learning in the low sample regime. You are invited to a Zoom meeting. When: Mar 26, 2020 03:00 PM Eastern Time (US and Canada) Host: Lenore Cowen

Register in advance for this meeting: https://tufts.zoom.us/meeting/register/vpUpceytqT4vhFFutlvgWmEnOzrOSl0pw

After registering, you will receive a confirmation email containing information about joining the meeting.